# US 12,118,706 B2 · Method and apparatus of evaluating quality of wafer or single crystal ingot > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,118,706 B2 - Document ID: US12118706B2 - Status: Granted - Kind: B2 - Application number: 17/647,388 - Filed: 2022-01-07 - Published: 2022-07-28 - Granted: 2024-10-15 - Earliest claimed: 2021-01-28 - Assignees: SK Siltron Co., Ltd. - Applicants: SK Siltron Co., Ltd. - Inventors: Hyun Woo Park - Web page: https://patentgrep.com/patent/US12118706B2 ## CPC - G06T 7/0004: {Industrial image inspection} - G01N 21/9505: {Wafer internal defects, e.g. microcracks} - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/08: Learning methods - G06N 3/09: Supervised learning - G06N 5/022: Knowledge engineering; Knowledge acquisition - G06T 3/00: Geometric image transformations in the plane of the image - G06T 7/11: Region-based segmentation - G01N 2021/8477: {Investigating crystals, e.g. liquid crystals} - G01N 2021/8883: {involving the calculation of gauges, generating models} - G06T 2207/20081: Training; Learning - G06T 2207/20084: Artificial neural networks [ANN] - G06T 2207/30148: Semiconductor; IC; Wafer - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A method of evaluating quality of a wafer or an apparatus of evaluating quality of a wafer may include: performing a copper-haze evaluation on a piece of wafer or a single crystal ingot; collecting copper-haze map data and a copper-haze evaluation score based on a result of the copper-haze evaluation; training an artificial intelligence model based on the copper-haze map data and the copper-haze evaluation score; and performing crystal defect evaluation on the piece of the wafer or the single crystal ingot using the learned artificial intelligence model that outputs the copper-haze evaluation score when the copper-haze map data is input. ## Claims ### Claim 1 A method of evaluating quality of a wafer, the method comprising: performing a copper-haze evaluation on a piece of wafer or a single crystal ingot; collecting copper-haze map data and a copper-haze evaluation score based on a result of the copper-haze evaluation; training an artificial intelligence model based on the copper-haze map data and the copper-haze evaluation score; and performing crystal defect evaluation on the piece of the wafer or the single crystal ingot using the learned artificial intelligence model that outputs the copper-haze evaluation score when the copper-haze map data is input, wherein the training of the artificial intelligence model based on the copper-haze map data and the copper-haze evaluation score comprises; performing learning data augmentation that generates at least one piece of virtual copper-haze map data based on the copper-haze map data; and training the artificial intelligence model using the copper-haze evaluation score corresponding to the virtual copper-haze map data generated according to the copper-haze map data and the learning data augmentation, and wherein the performing of the learning data augmentation comprises; dividing the copper-haze map data into a first region and a second region; performing image transformation on each of the divided first and second regions; and generating the virtual copper-haze map data based on each of the divided first and second regions on which the image transformation has been performed, wherein the first region and the second region are formed through a first heat treatment method and a second heat treatment method that are different from each other, respectively. ### Claim 2 (depends on 1) The method of claim 1, wherein the dividing of the copper-haze map data into the first region and the second region further comprises dividing each of the first region and the second region into an edge region and a center region. ### Claim 3 (depends on 1) The method of claim 1, wherein the image transformation includes at least one transformation of tilt, shearing, reflection, rotation, and clipping of an image. ### Claim 4 (depends on 1) The method of claim 1, wherein the performing of the crystal defect evaluation on the piece of the wafer or the single crystal ingot using the learned artificial intelligence model comprises: dividing each of the copper-haze map and virtual copper-haze map data into a first region and a second region using the learned artificial intelligence model; and calculating copper-haze scores for an edge region and a center region of each of the first region and the second region, respectively using the learned artificial intelligence model. ### Claim 5 An apparatus of evaluating quality of a wafer, the apparatus comprising: an input unit that receives copper-haze map data of a wafer; a processor that evaluates the quality of the wafer using an artificial intelligence model learned so as to output a copper-haze score when the copper-haze map data is input; and a display unit that outputs a result of the quality of the wafer, wherein the processor performs learning data augmentation that generates at least one piece of virtual copper-haze map data based on the copper-haze map data and trains the artificial intelligence model using a copper-haze evaluation score corresponding to the virtual copper-haze map data generated according to the copper-haze map data and the learning data augmentation, and wherein the processor divides the copper-haze map data into a first region and a second region, performs image transformation on each of the divided first and second regions, and generates the virtual copper-haze map data based on each of the divided first and second regions on which the image transformation is performed. ### Claim 6 (depends on 5) The apparatus of claim 5, wherein the processor further divides each of the first region and the second region into an edge region and a center region, performs the image transformation on each of the divided first and second regions and generates the virtual copper-haze map data based on each of the divided first and second regions on which the image transformation is performed. ### Claim 7 (depends on 5) The apparatus of claim 5, wherein the image transformation includes at least one transformation of tilt, shearing, reflection, rotation, and clipping of an image.